Arch Linux

andre.vmatos

Member

Registered: 2010-07-28

Posts: 19

Re: Discussion: the ati/radeon open source drivers & radeon repository

Unfortunatelly, here the problem persists. The kwin log when activate the compositing:

[andre@midichlorian ~ $] kwin --replace
kwin(6343) KWin::CompositingPrefs::detectDriverAndVersion: GL vendor is "Advanced Micro
Devices, Inc."
kwin(6343) KWin::CompositingPrefs::detectDriverAndVersion: GL renderer is "Mesa DRI R600
 (RS780 9612) 20090101  TCL DRI2"
kwin(6343) KWin::CompositingPrefs::detectDriverAndVersion: GL version is "2.0 Mesa 7.9-d
evel"
kwin(6343) KWin::CompositingPrefs::detectDriverAndVersion: Detected driver "radeon" , ve
rsion "20090101"
Application::crashHandler() called with signal 11; recent crashes: 1
KCrash: Application 'kwin' crashing...
KCrash: Attempting to start /usr/lib/kde4/libexec/drkonqi from kdeinit
sock_file=/home/andre/.kde4/socket-midichlorian/kdeinit4__0
X Error: BadWindow (invalid Window parameter) 3
  Major opcode: 20 (X_GetProperty)
  Resource id:  0x6a00035
kwin(6685) KWin::CompositingPrefs::detectDriverAndVersion: GL vendor is "Advanced Micro
Devices, Inc."
kwin(6685) KWin::CompositingPrefs::detectDriverAndVersion: GL renderer is "Mesa DRI R600
 (RS780 9612) 20090101  TCL DRI2"
kwin(6685) KWin::CompositingPrefs::detectDriverAndVersion: GL version is "2.0 Mesa 7.9-d
evel"
kwin(6685) KWin::CompositingPrefs::detectDriverAndVersion: Detected driver "radeon" , ve
rsion "20090101"
Application::crashHandler() called with signal 11; recent crashes: 2

And the packages:

[root@midichlorian ~ #] pacman -Sl radeon
radeon dri2proto-git 20100712-1 [instalado]
radeon glproto-git 20100712-1 [instalado]
radeon kernel26-drm-radeon-testing 20100721-1 [instalado]
radeon lib32-libdrm-git 20100721-1
radeon lib32-mesa-full 20100721-1
radeon libdrm-git 20100716-1 [instalado]
radeon libpthread-stubs 0.3-4 [instalado]
radeon libtxc_dxtn 070518-1 [instalado]
radeon linux-firmware-git 20100712-1 [instalado]
radeon mesa-demos-git 20100712-1 [instalado]
radeon mesa-full 20100728-1 [instalado]
radeon radeon_ucode 2-2
radeon xf86-video-ati-git 20100728-1 [instalado]

andre.vmatos

Member

Registered: 2010-07-28

Posts: 19

Re: Discussion: the ati/radeon open source drivers & radeon repository

Hi, thanx for the help. The composite with the opensource drivers worked here with KDE 4.4.5. Kwin of the KDE 4.5 RC3 doesn't work, crashing with the logs above.

korpenkraxar

Member

Registered: 2006-04-02

Posts: 123

Re: Discussion: the ati/radeon open source drivers & radeon repository

I have been compiling and using several 2.6.35 kernels in addition to kernel26-drm-radeon-testing and they all cause problems on my Thinkpad W500 with an ATI chip. X crashes at random occasions (using KDE with no effects) and I often just get a blinking cursor in the top-left corner when resuming from suspend. Do these problem currently occur for others as well?

Perry3D

Member

Registered: 2008-03-20

Posts: 553

Re: Discussion: the ati/radeon open source drivers & radeon repository

Hi,

on my system the current driver works stable. There was no crash for a very long time.
Do you get a hint in a log file? Is this just a freeze of X or a real kernel panic (does the sysrq keys work or can you log in with ssh?).

Btw: i would like to update the radeon repo, but i have some problems with the webspace and the maintainer (gtklocker) is on vacation .
Color tiling is now in master.

korpenkraxar

Member

Registered: 2006-04-02

Posts: 123

Re: Discussion: the ati/radeon open source drivers & radeon repository

Well, it could be related to frequently toggling between using no to several different external monitors, in my case over a Displayport->DVI connection. If I put the laptop too sleep before with one particular monitor connected and it woke up with a different monitor connected that did not support the particular mode, I would get a system lockup with the laptop screen going white. This particular crash behavior seems to have been fixed in the more recent kernels, but I would not be surprised if there is still remnants of my niche problem lurking around.

In any case, I have either had X die on me and be restarted by KDM, or X die and me being able to toggle among consoles and restart X myself or just toggle for a few seconds before the whole system dies. The latter may be related to toggling to console 7 where X was located before. I have not hunted around among the logs yet. Any suggestion which logs may be of particular use here?

gtklocker

Member

Registered: 2009-09-01

Posts: 462

Re: Discussion: the ati/radeon open source drivers & radeon repository

Btw: i would like to update the radeon repo, but i have some problems with the webspace and the maintainer (gtklocker) is on vacation

I replied you, man, didn't you see the email?

(Another Arch user which uploads to the same server has a problem (server us wrongly reporting a file is uploaded when it's not, so the upload can't continue and it asks to replace the [empty] file), so I'm waiting for the server owner to find out what's wrong )

gtklocker

Member

Registered: 2009-09-01

Posts: 462

Re: Discussion: the ati/radeon open source drivers & radeon repository

Ok, things are OK now. Perry please upload.

handy

Member

From: Oz

Registered: 2008-03-26

Posts: 719

Re: Discussion: the ati/radeon open source drivers & radeon repository

I primarily look to this thread to be informed of just what is happening in the ATi open source world; great posts, links & all.

Since the hosting of Perry's files has been down lately my source of information has never been so dry.

All of which has been quite an education for me, as I hadn't realised just how much I relied on this thread before.

Anyway, thanks to all of you guys that make this thread; past, present & future. I won't pass on any appreciation to Perry 'cause he's too humble to accept it anyway.

---

I used to be surprised that I was still surprised by my own stupidity, finding it strangely refreshing.
Well, now I don't find it refreshing.
I'm over it!

gee

Member

Registered: 2006-11-29

Posts: 313

Re: Discussion: the ati/radeon open source drivers & radeon repository

Hey everyone,

with the latest stack, was anyone able to get 5.1 setup through HDMI?
I can get passthrough but not 5.1...

I am using ALSA for the sound system, and whatever I do (even playing with asound.conf) speaker-test never outputs on more than 2.1...

Thanks

Perry3D

Member

Registered: 2008-03-20

Posts: 553

Re: Discussion: the ati/radeon open source drivers & radeon repository

He, i updated the kernel for the x86_64 repo (i686 is coming soon). But i experience a small bug with the new underscan feature for the tv output: it automatically applies to my DVI output. If someone has the same problem you can turn it off with:

xrandr --output DVI-0 --set underscan off

And i am thinking about applying the Responsiveness Patches: Phoronix: The Linux Desktop Responsiveness Patches Are Feeling Good

Perry3D

Member

Registered: 2008-03-20

Posts: 553

Re: Discussion: the ati/radeon open source drivers & radeon repository

Ok, i applied the responsiveness patch to the kernel. I love it .
I hope Linus will accept this patch.

megawebmaster

Member

Registered: 2009-08-12

Posts: 75

Re: Discussion: the ati/radeon open source drivers & radeon repository

Hmmm, anybody has random freezes of screen on the latest xf86-video-ati-git and drm-radeon-testing kernel from radeon repo? My notebook (with Mobility Radeon HD4570) makes screen blank, but music (from MPD) is still playing. I can't do anything but hard reset (I didn't try SysReq combination). It's quite annoying, currently I'm running stock kernel, but it's not what tigers loves

gtklocker

Member

Registered: 2009-09-01

Posts: 462

Re: Discussion: the ati/radeon open source drivers & radeon repository

Ok, i applied the responsiveness patch to the kernel. I love it .
I hope Linus will accept this patch.

Well, I can't see any difference.

BigLouis87

Member

From: Italy

Registered: 2010-05-15

Posts: 44

Re: Discussion: the ati/radeon open source drivers & radeon repository

Ok, i applied the responsiveness patch to the kernel. I love it .
I hope Linus will accept this patch.

Well, I can't see any difference.

I have read the Phoronics article and I have understood the problem is when there is a big load on disk, and the entire system freezes (even the mouse pointer!) for a little while... has this ever happened to you? Lots of times to me...
We will see in the future, when another block should happens....

PS: now I have realized my English is soooo bad! xD

gtklocker

Member

Registered: 2009-09-01

Posts: 462

Re: Discussion: the ati/radeon open source drivers & radeon repository

No, I haven't any freezes so far and I've updated to the new kernel since morning.

PS: Your English is good

megawebmaster

Member

Registered: 2009-08-12

Posts: 75

Re: Discussion: the ati/radeon open source drivers & radeon repository

So that's weird. From booting to stock kernel everything seems fine (apart from not reading thermal sensor on graphics card and some other improvements). This is something about this new responsiveness patch...

Perry3D

Member

Registered: 2008-03-20

Posts: 553

Re: Discussion: the ati/radeon open source drivers & radeon repository

I also get a gpu lockup if i use xbmc. But it's not the patch: i still get it when i remove the patch:

Aug 14 18:37:52 perry64 kernel: ------------[ cut here ]------------
Aug 14 18:37:52 perry64 kernel: WARNING: at drivers/gpu/drm/radeon/radeon_fence.c:239 radeon_fence_wait+0x3a6/0x400 [radeon]()
Aug 14 18:37:52 perry64 kernel: Hardware name: To Be Filled By O.E.M.
Aug 14 18:37:52 perry64 kernel: GPU lockup (waiting for 0x0000101E last fence id 0x0000101C)
Aug 14 18:37:52 perry64 kernel: Modules linked in: fuse ipv6 k10temp it87 hwmon_vid btrfs zlib_deflate crc32c libcrc32c snd_hda_codec_atihdmi snd_hda_codec_realtek tuner_simple tuner_types cpufreq_ondemand powernow_k8 freq_table mperf tuner snd_usb_audio snd_usbmidi_lib snd_intel8x0 tvaudio ir_sony_decoder tda7432 ir_jvc_decoder snd_hda_intel snd_seq_dummy msp3400 ir_rc6_decoder snd_hda_codec ir_rc5_decoder snd_seq_oss firewire_ohci firewire_core snd_seq_midi_event ir_nec_decoder crc_itu_t snd_seq bttv v4l2_common videodev v4l1_compat v4l2_compat_ioctl32 videobuf_dma_sg videobuf_core btcx_risc ir_common ir_core tveeprom snd_emu10k1 snd_rawmidi snd_ac97_codec snd_util_mem snd_hwdep snd_seq_device snd_pcm_oss snd_mixer_oss snd_bt87x snd_pcm snd_timer snd_page_alloc snd soundcore ac97_bus sg tg3 libphy irtty_sir sir_dev irda crc_ccitt edac_core edac_mce_amd i2c_piix4 evdev thermal wmi button processor slhc forcedeth rtc_cmos rtc_core rtc_lib ext4 mbcache jbd2 crc16 usbhid hid ohci_hcd radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core ehci_hcd usbcore sata_nv ata_generic pata_amd sr_mod pata_acpi cdrom sd_mod ahci libahci libata scsi_mod
Aug 14 18:37:52 perry64 kernel: Pid: 6867, comm: xbmc.bin Not tainted 2.6.35-11612-g78d5edf-dirty #1
Aug 14 18:37:52 perry64 kernel: Call Trace:
Aug 14 18:37:52 perry64 kernel: [<ffffffff8104815b>] ? warn_slowpath_common+0x7b/0xc0
Aug 14 18:37:52 perry64 kernel: [<ffffffff81048255>] ? warn_slowpath_fmt+0x45/0x50
Aug 14 18:37:52 perry64 kernel: [<ffffffffa017d2e6>] ? radeon_fence_wait+0x3a6/0x400 [radeon]
Aug 14 18:37:52 perry64 kernel: [<ffffffff810609c0>] ? autoremove_wake_function+0x0/0x30
Aug 14 18:37:52 perry64 kernel: [<ffffffffa0100e9d>] ? ttm_bo_wait+0x10d/0x1c0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa010413a>] ? ttm_bo_move_accel_cleanup+0xaa/0x2a0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa017dd12>] ? radeon_move_blit+0x142/0x1a0 [radeon]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa017e145>] ? radeon_bo_move+0xb5/0x190 [radeon]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa010069c>] ? ttm_tt_bind+0x6c/0xa0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa01012fb>] ? ttm_bo_handle_move_mem+0x14b/0x3d0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa01032d1>] ? ttm_bo_evict+0x171/0x3a0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa00db3b7>] ? drm_mm_split_at_start+0x17/0x70 [drm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa0102728>] ? ttm_mem_evict_first+0x148/0x1e0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa0103090>] ? ttm_bo_mem_space+0x3f0/0x4c0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa01035f9>] ? ttm_bo_move_buffer+0xf9/0x160 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa01036ee>] ? ttm_bo_validate+0x8e/0x130 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa017fa23>] ? radeon_bo_fault_reserve_notify+0xc3/0x100 [radeon]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa0104c32>] ? ttm_bo_vm_fault+0x52/0x2f0 [ttm]
Aug 14 18:37:52 perry64 kernel: [<ffffffffa017db6c>] ? radeon_ttm_fault+0x4c/0x70 [radeon]
Aug 14 18:37:52 perry64 kernel: [<ffffffff810bd3d2>] ? __do_fault+0x52/0x550
Aug 14 18:37:52 perry64 kernel: [<ffffffff810c043a>] ? handle_mm_fault+0x1da/0xa40
Aug 14 18:37:52 perry64 kernel: [<ffffffffa0193840>] ? radeon_gem_wait_idle_ioctl+0x0/0x140 [radeon]
Aug 14 18:37:52 perry64 kernel: [<ffffffff81182168>] ? prio_tree_insert+0x148/0x2a0
Aug 14 18:37:52 perry64 kernel: [<ffffffff810b9848>] ? vma_prio_tree_insert+0x28/0x60
Aug 14 18:37:52 perry64 kernel: [<ffffffff8103174e>] ? do_page_fault+0x15e/0x3d0
Aug 14 18:37:52 perry64 kernel: [<ffffffff810f1f69>] ? vfs_ioctl+0x39/0xc0
Aug 14 18:37:52 perry64 kernel: [<ffffffff810f24cd>] ? do_vfs_ioctl+0x7d/0x500
Aug 14 18:37:52 perry64 kernel: [<ffffffff812fac25>] ? page_fault+0x25/0x30
Aug 14 18:37:52 perry64 kernel: ---[ end trace 0fb18b12d235e60f ]---
Aug 14 18:37:52 perry64 kernel: [drm] Disabling audio support
Aug 14 18:38:05 perry64 ntpd[6183]: adjusting local clock by -0.188640s
Aug 14 18:38:11 perry64 kernel: CE: hpet increased min_delta_ns to 7500 nsec

Perry3D

Member

Registered: 2008-03-20

Posts: 553

Re: Discussion: the ati/radeon open source drivers & radeon repository

It could be this bug: https://bugs.freedesktop.org/show_bug.cgi?id=29556
I will try the patch mentioned from Marc.

Perry3D

Member

Registered: 2008-03-20

Posts: 553

Re: Discussion: the ati/radeon open source drivers & radeon repository

It seems that this patch fix the bug. I already uploaded the patched x86_64 kernel. I686 is coming in a few minutes.

handy

Member

From: Oz

Registered: 2008-03-26

Posts: 719

Re: Discussion: the ati/radeon open source drivers & radeon repository

I just installed the new files Perry. When the iMac boots it gets to the point where it early starts KMS & goes to a black screen & stays there?

I've commented out the radeon-testing kernel for the time being. I have things to do on the machine for some days so I can't do much in the way of testing at the moment.

---

I used to be surprised that I was still surprised by my own stupidity, finding it strangely refreshing.
Well, now I don't find it refreshing.
I'm over it!

Dejan

Member

Registered: 2009-08-25

Posts: 5

Re: Discussion: the ati/radeon open source drivers & radeon repository

Can someone tell me when we can expect 3D support for Evergreen?

roughL

Member

Registered: 2010-03-16

Posts: 7

Re: Discussion: the ati/radeon open source drivers & radeon repository

Can someone tell me when we can expect 3D support for Evergreen?

According to this phronix post which is quite old, 3D support for evergreen should be in mesa 7.9 which will be released by the end of this quarter. Also have a look at RadeonFeature from the x.org wiki to track the progress of the driver.

Vi0L0

Member

From: Poland

Registered: 2009-06-24

Posts: 1,349

Website

Re: Discussion: the ati/radeon open source drivers & radeon repository

There's finally open-source 2D/3D/video acceleration for ATI's Radeon HD 5000 "Evergreen" family of graphics processors

The ATI Evergreen Mesa Code Has Now Landed

Open-Source 2D, 3D For ATI Radeon HD 5000 Series GPUs

---

GPG key: http://pgp.mit.edu:11371/pks/lookup?op= … 2D653C3094  or  http://keys.gnupg.net/pks/lookup?op=get … 2D653C3094

handy

Member

From: Oz

Registered: 2008-03-26

Posts: 719

Re: Discussion: the ati/radeon open source drivers & radeon repository

Thanks ViOLO, a lot of people are going to be happy to hear that news.

---

I used to be surprised that I was still surprised by my own stupidity, finding it strangely refreshing.
Well, now I don't find it refreshing.
I'm over it!

LukynZ

Member

From: Ostrava, Czech Republic

Registered: 2008-11-13

Posts: 122

Re: Discussion: the ati/radeon open source drivers & radeon repository

Hello. I am using catalyst for some time, so I am curious, if there is some 3D performance progress against catalyst or catalyst is still only choice